Local switching systems process originating and terminating calls between customer lines connected to the switching system and trunks to other switching systems. Since most customer lines are connected only to a single local switching system, the availability of telecommunication service to a customer is dependent upon the continued operability of the local switching system.
Such systems are designed and constructed to provide continuing reliable service in the presence of faults. The reliability is sometimes achieved by providing duplicate (redundant) equipment in the more critical parts of the system. When a failing part is detected, the duplicate is placed in service and the failing part is removed. With care, the substitution of duplicate for failing parts can occur without the loss of service to customers.
A distributed switching system is one comprised of a plurality of switch units which interface customer lines and trunks on a peripheral side of the unit and which interface a central switch arrangement on the other side of the unit. Calls between customers and/or trunks connected to different switch units are completed through the central switch. A duplicated distributed switching system comprises a plurality of duplicated switching units which are switchably connected by a duplicated pair of central switches. In the presence of faults, a duplicate switching unit is substituted for a failing one and a duplicate central switch is substituted when a central switch fails. This arrangement provides excellent service for systems which undergo normal failures in only one unit of the duplicated pairs.
A disaster, such as a fire in a switching office, may cause catastrophic failures which are not normal and result in the loss of both units of a duplicated pair. With no duplicate to substitute for a failing unit, the system may not be able to complete calls. For example, the failure of both units of the central switch will isolate the customers connected to one switch unit from the lines and trunks connected to all of the other switch units of the system. Not only is it impossible to reach a large number of the customers on the same switching system, a customer's access to other switching systems normally reached via trunks at other switching units is also lost. Such isolation of customers is a substantial service limitation which is compounded by the fact that damage repair after a disaster may take an extended period of time.
Although the subscriber isolation problem is raised in the context of disaster which destroys the central switch, the substantial isolation of customers can occur in less catastrophic ways. The switch units of a distributed switching system are connected to the central switch by communication links. The complete failure of the communication link between a switch unit and the central switch can isolate the customers of the switching unit served by that link as much as the complete central switch failure.
The co-pending application of Cline et al., case 1-1-1 provides inter-switch unit trunks as an alternative to the central switch communication paths between switching units. Inter-switch unit trunks provide a valuable connection resource when "normal" connections through the central switch are unavailable. However, the inter-switch unit trunks require trunk services such as scanning and signaling which use valuable switch unit controller resources. Also the switch units must provide the capacity to connect inter-switch unit trunks through the switch unit. The controller resources required to serve the inter-switch unit trunks and the connection resources required for connections using the inter-switch unit trunks burden each switch unit and reduce the number of lines and inter-office trunks that a switch unit can serve. Since inter-switch unit trunks are used only when duplicate unit failures occur, the probability of using the inter-switch unit trunks is low and the constant overhead needed to serve the inter-switch unit trunks is not desirable.
Accordingly, a need exists in the art for a telecommunication switching system which can provide continuing switching unit-to-switching unit connections when the central switching arrangement of a distributed switching system cannot be employed to communicate from one switch unit to another and which does not substantially increase the "normal" operation overhead of the switch units.